[DO NOT MERGE] Numpy2

.github/workflows/build.yaml

@@ -72,7 +72,7 @@ jobs:
 
       - uses: actions/setup-python@v5
         with:
-          python-version: '3.11'
+          python-version: '3.12'
 
       - name: Build sdist
         run: pipx run build --sdist

.github/workflows/test.yaml

@@ -20,22 +20,23 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
-      - name: Set up Python 3.11
+      - name: Set up Python 3.12
         uses: actions/setup-python@v5
         with:
-          python-version: 3.11
+          python-version: 3.12
       - name: Linting
         run: |
           pip install pre-commit
           pre-commit run --all-files
 
   test:
-    name: Test on ${{ matrix.os }}, Python ${{ matrix.python-version }}
+    name: Test on ${{ matrix.os }}, Python ${{ matrix.python-version }}, Numpy ${{ matrix.numpy-version }}
     runs-on: "${{ matrix.os }}"
     strategy:
       matrix:
         os: ["ubuntu-latest", "macos-13", "macos-latest"]  # macos-13 is x86-64, macos-latest is arm64
         python-version: ["3.9", "3.10", "3.11", "3.12"]
+        numpy-version: ['1', '2']
 
     steps:
     - uses: actions/checkout@v4
@@ -53,7 +54,7 @@ jobs:
     - uses: conda-incubator/setup-miniconda@v3
       with:
         python-version: ${{ matrix.python-version }}
-        environment-file: devtools/conda-envs/test_env.yaml
+        environment-file: devtools/conda-envs/test_env_numpy${{ matrix.numpy-version }}.yaml
         activate-environment: test_env
         channel-priority: true
         auto-update-conda: true
@@ -64,6 +65,7 @@ jobs:
       # conda setup requires this special shell
       shell: bash -l {0}
       run: |
+        conda list
         conda info --all
         python -m pip install . -v --no-deps
         conda list
@@ -96,7 +98,7 @@ jobs:
       - uses: "actions/checkout@v4"
       - uses: "actions/setup-python@v5"
         with:
-          python-version: "3.11"
+          python-version: "3.12"
       - name: "Install HDF5 with brew if macos-arm64"
         if: ${{ matrix.os == 'macos-latest' }}
         run: "brew install hdf5"

devtools/conda-envs/test_env.yaml

@@ -5,7 +5,7 @@ channels:
 dependencies:
   - python
   - cython
-  - numpy<2
+  - numpy
   - scipy
   - h5py
   - pyyaml

devtools/conda-envs/test_env_numpy1.yaml

@@ -0,0 +1,25 @@
+name: test_env
+channels:
+  - conda-forge
+  - defaults
+dependencies:
+  - python
+  - cython
+  - numpy=1
+  - scipy
+  - h5py
+  - pyyaml
+  - pyzmq
+  - matplotlib-base
+  - blessings
+  - ipykernel
+  - mpi4py
+  - tqdm
+  - mdtraj
+  - netCDF4
+  # testing
+  - pytest
+  - pytest-cov
+  - pytest-rerunfailures
+  - pytest-timeout
+  - codecov

devtools/conda-envs/test_env_numpy2.yaml

@@ -0,0 +1,24 @@
+name: test_env
+channels:
+  - conda-forge
+  - defaults
+dependencies:
+  - python
+  - cython
+  - numpy
+  - scipy
+  - h5py
+  - pyyaml
+  - pyzmq
+  - matplotlib-base
+  - blessings
+  - ipykernel
+  - mpi4py
+  - tqdm
+  - mdtraj
+  # testing
+  - pytest
+  - pytest-cov
+  - pytest-rerunfailures
+  - pytest-timeout
+  - codecov

pyproject.toml

@@ -7,8 +7,7 @@ requires = [
     "wheel",
     "Cython >=0.29.16; python_version >='3.10'",  # Note: sync with setup.py
     "Cython <3.0.3, >=0.29.16; python_version <'3.10'",
-    "oldest-supported-numpy; python_version <'3.9'",
-    "numpy >=1.25.0, <2; python_version >='3.9'",
+    "numpy >=2.0.0, <3",  # Force numpy>=2 during build time so it works on both 1.XX and 2.XX
     "scipy >=0.19.1",
     "versioneer-518"
 ]
@@ -45,3 +44,6 @@ exclude = '''
      )
 )
 '''
+
+[tool.ruff.lint]
+select = ['NPY201']

setup.py

@@ -109,7 +109,7 @@ def extensions():
 ]
 
 INSTALL_REQUIRES = [
-    "numpy >= 1.16.0, <2",
+    "numpy >= 1.25, <3",
     "scipy >= 0.19.1",
     "h5py >= 2.10",
     "mdtraj >= 1.9.5",
@@ -139,7 +139,7 @@ def extensions():
     version=versioneer.get_version(),
     keywords='',
     cmdclass=versioneer.get_cmdclass(),
-    python_requires=">=3.6",
+    python_requires=">=3.9",
     zip_safe=False,
     classifiers=CLASSIFIERS,
     entry_points={'console_scripts': console_scripts},

src/westpa/cli/tools/w_assign.py

@@ -474,17 +474,17 @@ def go(self):
 
             # Recursive mappers produce a generator rather than a list of labels
             # so consume the entire generator into a list
-            labels = [np.string_(label) for label in self.binning.mapper.labels]
+            labels = [np.bytes_(label) for label in self.binning.mapper.labels]
 
             self.output_file.create_dataset('bin_labels', data=labels, compression=9)
 
             if self.states:
                 nstates = len(self.states)
                 state_map[:] = nstates  # state_id == nstates => unknown state
-                state_labels = [np.string_(state['label']) for state in self.states]
+                state_labels = [np.bytes_(state['label']) for state in self.states]
 
                 for istate, sdict in enumerate(self.states):
-                    assert state_labels[istate] == np.string_(sdict['label'])  # sanity check
+                    assert state_labels[istate] == np.bytes_(sdict['label'])  # sanity check
                     state_assignments = assign(sdict['coords'])
                     for assignment in state_assignments:
                         state_map[assignment] = istate
@@ -558,7 +558,7 @@ def go(self):
                 shuffle=True,
                 chunks=h5io.calc_chunksize(pops_shape, weight_dtype),
             )
-            h5io.label_axes(pops_ds, [np.string_(i) for i in ['iteration', 'state', 'bin']])
+            h5io.label_axes(pops_ds, [np.bytes_(i) for i in ['iteration', 'state', 'bin']])
 
             pi.new_operation('Assigning to bins', iter_stop - iter_start)
             last_labels = None  # mapping of seg_id to last macrostate inhabited

src/westpa/cli/tools/w_fluxanl.py

@@ -11,7 +11,7 @@
 from westpa.tools.dtypes import iter_block_ci_dtype as ci_dtype
 import westpa.mclib as mclib
 
-fluxentry_dtype = np.dtype([('n_iter', n_iter_dtype), ('flux', weight_dtype), ('count', np.uint)])
+fluxentry_dtype = np.dtype([('n_iter', n_iter_dtype), ('flux', weight_dtype), ('count', np.int32)])
 
 target_index_dtype = np.dtype(
     [

src/westpa/cli/tools/w_ntop.py

@@ -190,9 +190,9 @@ def go(self):
                         weights = all_weights.take(segs)
 
                         if what == 'lowweight':
-                            indices = np.argsort(weights)[:count]
+                            indices = np.argsort(weights, kind='stable')[:count]
                         elif what == 'highweight':
-                            indices = np.argsort(weights)[::-1][:count]
+                            indices = np.argsort(weights, kind='stable')[::-1][:count]
                         else:
                             assert what == 'random'
                             indices = np.random.permutation(len(weights))

src/westpa/cli/tools/w_red.py

@@ -3,6 +3,11 @@
 from westpa import rc
 from westpa.tools import WESTParallelTool
 
+try:
+    from numpy import trapezoid
+except ImportError:
+    from numpy import trapz as trapezoid
+
 
 class DurationCorrector(object):
     @staticmethod
@@ -128,9 +133,9 @@ def correction(self, iters, freqs=None):
 
         for i, tau in enumerate(taugrid):
             if i > 0 and tau < maxduration:
-                integral1[i] = np.trapz(f_tilde[: i + 1], taugrid[: i + 1])
+                integral1[i] = trapezoid(f_tilde[: i + 1], taugrid[: i + 1])
 
-        integral2 = np.trapz(integral1, taugrid)
+        integral2 = trapezoid(integral1, taugrid)
 
         if integral2 == 0:
             return 0.0

src/westpa/core/binning/assign.py

@@ -134,7 +134,7 @@ def boundaries(self, boundaries):
         self._boundaries = []
         self.labels = labels = []
         for boundset in boundaries:
-            boundarray = np.asarray(boundset, dtype=coord_dtype, order='C')
+            boundarray = np.ascontiguousarray(boundset, dtype=coord_dtype)
             db = np.diff(boundarray)
             if (db <= 0).any():
                 raise ValueError('boundary set must be strictly monotonically increasing')
@@ -146,8 +146,12 @@ def boundaries(self, boundaries):
         _boundaries = self._boundaries
         binspace_shape = tuple(self._boundlens[:] - 1)
         for index in np.ndindex(binspace_shape):
-            bounds = [(_boundaries[idim][index[idim]], boundaries[idim][index[idim] + 1]) for idim in range(len(_boundaries))]
-            labels.append(repr(bounds))
+            bounds = [
+                (float(str(_boundaries[idim][index[idim]])), float(str(boundaries[idim][index[idim] + 1])))
+                for idim in range(len(_boundaries))
+            ]
+            print(repr(bounds))
+            labels.append(str(bounds))
 
     def assign(self, coords, mask=None, output=None):
         try:
@@ -184,7 +188,7 @@ def __init__(self, functions):
         self.functions = functions
         self.nbins = len(functions)
         self.index_dtype = np.min_scalar_type(self.nbins)
-        self.labels = [repr(func) for func in functions]
+        self.labels = [str(func) for func in functions]
 
     def assign(self, coords, mask=None, output=None):
         if output is None:

src/westpa/core/binning/mab.py

@@ -173,8 +173,8 @@ def map_mab(coords: np.ndarray, mask: np.ndarray, output: List[int], *args, **kw
     if skip is None:
         skip = [0] * ndim
 
-    allcoords = np.copy(coords)
-    allmask = np.copy(mask)
+    allcoords = coords.copy()
+    allmask = mask.copy()
 
     weights = None
     isfinal = None
@@ -294,16 +294,20 @@ def detect_bottlenecks(unmasked_coords, unmasked_weights, n_coords, n):
     """
     # Grabbing all unmasked coords in current dimension, plus corresponding weights
     # Sort by current dimension in coord, smallest to largest
-    sorted_indices = unmasked_coords[:, n].argsort()
+    sorted_indices = unmasked_coords[:, n].argsort(kind='stable')
+
     # Grab sorted coords and weights
     coords_srt = unmasked_coords[sorted_indices, :]
     weights_srt = unmasked_weights[sorted_indices]
+
     # Also sort in reverse order for opposite direction
     coords_srt_flip = np.flipud(coords_srt)
     weights_srt_flip = np.flipud(weights_srt)
+
     # Initialize the max directional differences along current dimension as None (these may not be updated)
     bottleneck_coords, bottleneck_coords_flip = None, None
     maxdiff, maxdiff_flip = -np.inf, -np.inf
+
     # Looping through all non-boundary coords
     # Compute the cumulative weight on either side of each non-boundary walker
     for i in range(1, n_coords - 1):

src/westpa/core/data_manager.py

@@ -1549,7 +1549,7 @@ def create_dataset_from_dsopts(group, dsopts, shape=None, dtype=None, data=None,
         #        dsopts['file'] = str(dsopts['file']).format(n_iter=n_iter)
         h5_auxfile = h5io.WESTPAH5File(dsopts['file'].format(n_iter=n_iter))
         h5group = group
-        if not ("iter_" + str(n_iter).zfill(8)) in h5_auxfile:
+        if ("iter_" + str(n_iter).zfill(8)) not in h5_auxfile:
             h5_auxfile.create_group("iter_" + str(n_iter).zfill(8))
         group = h5_auxfile[('/' + "iter_" + str(n_iter).zfill(8))]
 
@@ -1652,7 +1652,7 @@ def create_dataset_from_dsopts(group, dsopts, shape=None, dtype=None, data=None,
     if 'file' in list(dsopts.keys()):
         import h5py
 
-        if not dsopts['h5path'] in h5group:
+        if dsopts['h5path'] not in h5group:
             h5group[dsopts['h5path']] = h5py.ExternalLink(
                 dsopts['file'].format(n_iter=n_iter), ("/" + "iter_" + str(n_iter).zfill(8) + "/" + dsopts['h5path'])
             )

src/westpa/core/h5io.py

@@ -344,10 +344,10 @@ def label_axes(h5object, labels, units=None):
     if len(units) and len(units) != len(labels):
         raise ValueError('number of units labels does not match number of axes')
 
-    h5object.attrs['axis_labels'] = np.array([np.string_(i) for i in labels])
+    h5object.attrs['axis_labels'] = np.array([np.bytes_(i) for i in labels])
 
     if len(units):
-        h5object.attrs['axis_units'] = np.array([np.string_(i) for i in units])
+        h5object.attrs['axis_units'] = np.array([np.bytes_(i) for i in units])
 
 
 NotGiven = object()

src/westpa/core/kinetics/_kinetics.pyx

@@ -8,7 +8,7 @@ ctypedef np.uint16_t index_t
 ctypedef np.float64_t weight_t
 ctypedef np.uint8_t bool_t
 ctypedef np.int64_t seg_id_t
-ctypedef np.uint_t uint_t # 32 bits on 32-bit systems, 64 bits on 64-bit systems
+ctypedef np.uintp_t uint_t # 32 bits on 32-bit systems, 64 bits on 64-bit systems
 
 cdef double NAN = np.nan
 

src/westpa/core/kinetics/matrates.py

@@ -61,7 +61,7 @@ def get_steady_state(rates):
 
     vals = np.abs(vals)
     log.debug('eigenvalues: {!r}'.format(list(reversed(sorted(vals)))))
-    asort = np.argsort(vals)
+    asort = np.argsort(vals, kind='stable')
     vec = vecs[:, asort[-1]]
     ss = np.abs(vec)
 

src/westpa/core/reweight/_reweight.pyx

@@ -21,7 +21,7 @@ ctypedef numpy.uint16_t index_t
 ctypedef numpy.float64_t weight_t
 ctypedef numpy.uint8_t bool_t
 ctypedef numpy.int64_t trans_t
-ctypedef numpy.uint_t uint_t # 32 bits on 32-bit systems, 64 bits on 64-bit systems
+ctypedef numpy.uintp_t uint_t # 32 bits on 32-bit systems, 64 bits on 64-bit systems
 ctypedef unsigned short Ushort
 ctypedef double complex Cdouble